lkml.org 
[lkml]   [2021]   [Nov]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: How to reduce PCI initialization from 5 s (1.5 s adding them to IOMMU groups)s
From
Dear Linux folks,


Am 05.11.21 um 12:56 schrieb Paul Menzel:

> On a PowerEdge T440/021KCD, BIOS 2.11.2 04/22/2021, Linux 5.10.70 takes
> almost five seconds to initialize PCI. According to the timestamps, 1.5
> s are from assigning the PCI devices to the 142 IOMMU groups.
>
> ```
> $ lspci | wc -l
> 281
> $ dmesg
> […]
> [    2.918411] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
> [    2.933841] ACPI: Enabled 5 GPEs in block 00 to 7F
> [    2.973739] ACPI: PCI Root Bridge [PC00] (domain 0000 [bus 00-16])
> [    2.980398] acpi PNP0A08:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI HPX-Type3]
> [    2.989457] acpi PNP0A08:00: _OSC: platform does not support [LTR]
> [    2.995451] acpi PNP0A08:00: _OSC: OS now controls [PME PCIeCapability]
> [    3.001394] acpi PNP0A08:00: FADT indicates ASPM is unsupported,
> using BIOS configuration
> [    3.010511] PCI host bridge to bus 0000:00
> […]
> [    6.233508] system 00:05: [io  0x1000-0x10fe] has been reserved
> [    6.239420] system 00:05: Plug and Play ACPI device, IDs PNP0c02 (active)
> [    6.239906] pnp: PnP ACPI: found 6 devices
> […]
> [    6.989016] pci 0000:d7:05.0: disabled boot interrupts on device [8086:2034]
> [    6.996063] PCI: CLS 0 bytes, default 64
> [    7.000008] Trying to unpack rootfs image as initramfs...
> [    7.065281] Freeing initrd memory: 5136K
> […]
> [    7.079098] DMAR: dmar7: Using Queued invalidation
> [    7.083983] pci 0000:00:00.0: Adding to iommu group 0
> […]
> [    8.537808] pci 0000:d7:17.1: Adding to iommu group 141
> [    8.571191] DMAR: Intel(R) Virtualization Technology for Directed I/O
> [    8.577618] PCI-DMA: Using software bounce buffering for IO (SWIOTLB)
> […]
> ```
>
> Is there anything that could be done to reduce the time?

I created an issue at the Kernel.org Bugzilla, and attached the output
of `dmesg` there [1].


Kind regards,

Paul


[1]: https://bugzilla.kernel.org/show_bug.cgi?id=214953

\
 
 \ /
  Last update: 2021-11-05 13:04    [W:0.688 / U:0.012 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site